|Publication number||US7253986 B2|
|Application number||US 11/215,602|
|Publication date||Aug 7, 2007|
|Filing date||Aug 30, 2005|
|Priority date||Aug 30, 2005|
|Also published as||CN1924999A, CN100405461C, US20070047131|
|Publication number||11215602, 215602, US 7253986 B2, US 7253986B2, US-B2-7253986, US7253986 B2, US7253986B2|
|Inventors||David Berman, W. Stanley Czarnecki, Icko F. Iben, Wayne Isami Imaino|
|Original Assignee||International Business Machines Corporation|
|Export Citation||BiBTeX, EndNote, RefMan|
|Patent Citations (14), Non-Patent Citations (23), Referenced by (29), Classifications (8), Legal Events (4)|
|External Links: USPTO, USPTO Assignment, Espacenet|
This application is related to U.S. patent application Ser. No. 11/216,615 to Czarnecki et al., entitled “Tape Head Having Write Devices and Narrower Read Devices”, filed concurrently herewith, and which is herein incorporated by reference.
The present invention relates to tape drive heads, and more particularly, this invention relates to a write and reader array where widths of the readers are smaller than widths of the writers or data tracks.
Data is stored on magnetic media such as tape by writing data in a multiplicity of linear tracks. The tracks are separated along the transverse direction of the tape and a given track runs longitudinally along the tape.
In an effort to increase the amount of data that can be written for a given tape width, efforts have been made to make data tracks adjacent to one another. The most common method for writing is to use writers that are spaced apart by a predetermined distance. Furthermore, the predominant method of writing is to have a large separation between readers and a large separation between writers. Adjacent tracks are written in separate passes of the tape where the head is stepped over in the horizontal or transverse direction by the desired track width. The writer width is wider than the desired track width. With each pass, the newly written track overlaps the previously written track so the resulting width of the previous track is the desired final track width. The above described method is termed “shingling”. Another method is to write adjacent tracks simultaneously. As the separation between tracks becomes narrower, horizontal motions of the writing/reading heads relative to the tape will reach values a fraction of the desired read/write track widths.
The technology used in existing tape storage drives aligns the readers within the width of a written track so each reader is aligned over a single track. The reader is typically smaller than the writer, is aligned therewith, and is reading one single track. This method is called “write wide, read narrow.” Because the reader is narrower than the writer, the reader will tend not to read adjacent tracks in spite of the horizontal “wobble” of the tape relative to the reader as the tape moves across the head.
A major drawback to the traditional “shingling” method, however, is that tape wobble increases the probability of overwriting adjacent data tracks during writing the reverse direction and also causes a random variation in the track width along the length of the tape. As the track width decreases, the amount of wobble (or track mis-registration) needs to decrease proportionally. As the track width is decreasing with future generations, it is becoming more difficult to decrease the track mis-registration sufficiently to keep readers on track and avoid overlap of readers on multiple written tracks.
One approach to control the written tracks is to use adjacent writers so a large group of adjacent tracks will be written simultaneously. Any horizontal motion (wobble) during write will cause the simultaneously-written tracks to move together, so the track-to-track separation (pitch) remains fixed within the group. Horizontal motion (wobble) results in large track misregistration during read, as a given reader can straddle two adjacent written tracks.
Furthermore, the servo tracks used for guiding the head-to-tape-track alignment are typically written to the tape prior to writing any data. Thus, any wobble of a written track will not be contained in the servo tracks. So, during readback, even though the head is following the servo tracks, errors occur due to the wobble during both writing and readback. The errors can result in a particular reader reading two or more tracks simultaneously, especially where track spacing is minimal. The resultant signal is a composition of two fields from both tracks and may make extraction of the data from any single track impossible.
One approach to solve these problems would be to use a multiplicity of writers and readers, where the number of readers is greater than the number of writers and to allow for the readers to be misaligned with respect to the written tracks so that each reader will have components of more than one track. The data would then be deconvoluted using an algorithm that took the interference into account. A major difficulty in the deconvolution is that the group of written tracks will wobble (or wander) in the horizontal location along the length of the tape.
There is accordingly a clearly-felt need in the art for a head assembly and method for accurately and efficiently deconvoluting a read signal reflecting multiple written data tracks, thereby allowing accurate reading of data in spite of tape wobble. These unresolved problems and deficiencies are clearly felt in the art and are solved by this invention in the manner described below.
To resolve the aforementioned problems, a system is provided for reading a magnetic medium having a data band thereon, the data band comprising a plurality of simultaneously written data tracks and at least one alignment band. The system in one embodiment includes a plurality of adjacent readers for simultaneously reading the data tracks. At least one reader is also present for reading the at least one alignment band. A mechanism determines a fractional overlap of each reader on the data tracks based on readback of the alignment band. Also, a mechanism extracts data from readback of the data tracks based at least in part on the fractional overlap.
Several methods for determining a fractional overlap of each reader on the data tracks based on readback of the alignment band and deconvolving the data from the data tracks are presented.
Any of these embodiments may be implemented in a tape drive system, which may include a magnetic head including the readers mentioned above, a drive mechanism for passing a magnetic recording tape over the magnetic head, and a controller electrically coupled to the magnetic head.
Other aspects and advantages of the present invention will become apparent from the following detailed description, which, when taken in conjunction with the drawings, illustrate by way of example the principles of the invention.
For a fuller understanding of the nature and advantages of the present invention, as well as the preferred mode of use, reference should be made to the following detailed description read in conjunction with the accompanying drawings.
The following description is the best embodiment presently contemplated for carrying out the present invention. This description is made for the purpose of illustrating the general principles of the present invention and is not meant to limit the inventive concepts claimed herein. Further, particular features described herein can be used in combination with other described features in each of the various possible combinations and permutations.
The following description discloses a method and system for successfully and accurately reading multiple data tracks where the readers overlap multiple tracks.
In order to increase the stability of the module 200 for the suitable use thereof, the module 200 is attached to a beam 206 of some sort formed of a rigid material. Such beams 206 are often referred to as a “U-beam.” A closure 208 is often attached in view of the benefits it affords in resultant heads.
The head 300 of
It should be noted that the two-module tape head 300 of
One skilled in the art will appreciate that the configuration of write and/or readers 201, 202 can vary. For instance, one module can have all writers 201, while the other module can have all readers 202. Another example would be to have a plurality of writers 201 and readers 202 all aligned linearly perpendicular to the direction of tape movement. It should also be understood that the number of read and writers described herein are provided by way of example only, and can be increased or decreased per the desires of the designer, system requirements and capabilities, etc.
Another variation includes a head having only a single module of read and writers that provides all of the read/write functionality. A second module may or may not be present. Of course the shape of the module may be different than the module 200 shown in
As shown, the writers are adjacent to one another, and the readers are adjacent to one another. Adjacent refers to the horizontal alignment. Each reader is horizontally located adjacent to its neighbor with little or no horizontal separation. The adjacent alignment can also be accomplished by displacement of neighboring readers in the vertical direction (direction of tape motion) for physical considerations such as avoiding overlapping leads used to connect the readers to external devices, etc. During reading, some readers overlapping two data tracks will generate a convoluted signal reflecting influence from the two data tracks. The read signals representing multiple tracks per reader can then be deconvoluted to extract the original information written on each individual data track. For the following discussion, d represents a fractional overlap of a reader over one data track and f represents the fractional overlap of the reader over another data track. If, for example, read track n (R(n)) has a fraction f of written track n (f*W(n)) and a fraction d of written track n+1(d*W(n+1)), then the vector read signal for readers R is described by the matrix M times the vector written signal W: R=M*W. In the simple case where the readers are homogeneous, and all readers behave the same, f and d are described by Equations 1 and 2 respectively, below. The diagonal of M is given by f (M(i,i)=f), and the only nonzero components of the off-diagonal elements would be M(i,i−1)=d. In other words, where the reader is only over one track, f will equal 1. Where the reader is overlapping two tracks, the signal generated by the reader will be dependent upon f and d. Because in this embodiment of the present invention the reader can only overlap two tracks, the only nonzero components of the off-diagonal elements would be d. If the values of f and d are not known, then the algorithm will need to determine both f and d to deconvolve the signals R to obtain the written signals W. A major difficulty in the deconvolution process is that the group of written tracks will wobble (or wander) in the horizontal location along the length of the tape, making the values of f and d change along the track length. As described below, this difficulty has been overcome by the present invention.
To enable the system to determine the relative position of the readers 202 relative to the data tracks (e.g., the overlap of the readers relative to the tracks) an alignment band is used. The alignment band is written concurrently with the data tracks. The alignment band can include one or more written tracks of a known pattern (control track(s)) on the medium. The alignment band can also merely be on one or more erase or “blank” tracks. Suggested alignment signals are described below, and generally include a combination of “blank” tracks and control tracks.
As shown in each case of
Because the tracks are always uniformly spaced, the fractional signals read by the readers overlapping the isolated control track will be identical to the fraction of signals from readers overlapping any other written tracks, unambiguously allowing determination of the matrix inversion. The width of the outer DC OR AC erase tracks is preferably at least one track width wide, but might be chosen to be larger depending on the amount of track misregistration. Complications such as those due to differences in the response signal of a reader or non-homogeneous response along a reader/writer track width can be determined by a calibration procedure and the values stored in a look up table. The inversion matrix can then be appropriately adjusted.
To illustrate the general method according to one embodiment of the present invention, assume N read tracks are on a head and P written tracks are present, with the width of the read and write tracks being nearly identical. Also assume N is greater than P, and that the writers are adjacent and the readers are adjacent. The signal for reader track n (R(n)) has a fraction f of written track n (f*W(n)) and a fraction d of written track n+1 (d*W(n+1)). (Assuming the signals have the same intensity and the readers abut one another, then d=1−f). As mentioned above, d represents a fractional overlap of a reader over one data track and f represents the fractional overlap of the reader over another data track.
f=x/W R Equation 1
d=y/W R=1−f Equation 2
The vector signal R for the readers is described by the matrix M times the vector signal W: R=M*W. The diagonal of M is given by f (M(i,i)=f), and the only nonzero components of the off-diagonal elements would be M(i,i−1)=d. If the values of f and d are not known, then the algorithm will need to determine both f and d to deconvolve the signals R to obtain the written signals W. A major difficulty in prior methods of deconvolution was that the group of written tracks will wobble (or wander) in the horizontal location along the length of the tape, making the values of f and d change along the track length. The writers used to write the tracks in this method of the present invention are aligned adjacent to one another and all tracks are written simultaneously so any wander in the bundle of tracks will be identical for all written tracks. The resulting tracks will be adjacent to one another and have about a constant (though not necessarily equal) center to center spacing along the length of the band of tracks.
For simplicity, and to match the illustrations in
The inverse of matrix M is:
The written track vector is: W=(W(1), W(2), W(3), W(4), 0)
while the read vector is: R=(R(1), R(2), R(3), R(4), R(5)).
The general solution to the inversion matrix for the number of written tracks being Ntrack requires a matrix dimension of Ndim=Ntrack+1, and assuming equal amplitude response for all readers and a uniform response along each reader track width.
The read signal given the written signal will be:
R=M*W Equation 3
The deconvolved written tracks (DW) are given by:
DW=IM*R=IM*M*W Equation 4
An equally valid matrix is:
The inverse of matrix M2 is:
The written track vector is: W2=(0, W(1), W(2), W(3), W(4)) while the read vector is: R2=(R(1), R(2), R(3), R(4), R(5)).
This second matrix (IM2) would be a better choice when d>f, especially when f<<1, due to errors in the signal to noise ratio (SNR) and the 1/f factors. The first matrix (IM) would be better when f>d. The two can also be combined to offer better SNR.
If an additional, isolated written track(s) WE(j) are made simultaneously (assuming at least one additional “extra” control track is written, surrounded by “blank” zones), the horizontal motion of the extra track(s) will be identical to that of the tracks with the “random” information written on the data tracks W. Since the extra track(s) is isolated (due to the “blank” zones, e.g., DC OR AC erased), the reader reading that signal will only get the signal from that track. With one extra control track (WE(1)), the values of f and d can be determined by Equations 5 and 6 (refer also to FIG. 7).
f=x/W R =RE(1)/[RE(1)+RE(2)] Equation 5
d=y/W R=1−f=RE(2)/[RE(1)+RE(2)] Equation 6
When the width of the readers is less than the width of the writers so a spacing exists between readers, f and d are still given by the above two equations.
“Dead” regions (e.g., drop out zones) on the tape might occur where the extra written track is not written, so having an extra written track on either end of the group of data tracks might be desired and would give the user more flexibility in determining the relative off-track coupling.
For a tape to which no data has been written, the “blank” zone regions will inherently exist simply by not writing data over that region. Once a tape has been written to, transverse motion of the tape with respect to the head results in a variation in the location of the “blank” zone region for different passes over the same region of the tape. This may necessitate the creation of a “blank” zone whenever the tape is written to. This can be accomplished by having writers over the “blank” zone regions erase the tape in the “blank” zone simultaneously to data being written (Wb(1)-Wb(3)). This will ensure that the “blank” zones exist and follow the “wobble” of the written data tracks. An example of a tape erasure would be a “DC” erasure performed by writers over tracks Wb(1)-Wb(4) being powered with a constant current throughout the entire write process, sufficient to magnetizing the tape in the “blank zone regions in one orientation without any transitions. An “AC” erasure generally refers to applying a sufficiently high frequency current to the writers at a sufficiently high current level to write alternately oriented magnetic transitions on the tape at a physical spacing small enough that the read heads could not read them.
Servos and servo tracks may be used to maintain track following (and may be considered “control” tracks), but extra written control track(s) along with the “blank” zones will greatly assist in deconvoluting the signals due to overlap of the different written tracks onto each reader. The extra written tracks can also serve as a fine tune servo signal. The inversion matrix (IM) is uniquely determined with only the knowledge of f and d. Application of the inversion matrix is a simple summation and multiplication. While the extra written tracks and the “blank zone” separating the extra written tracks from the data tracks uses storage space, with a large number of simultaneously written tracks (16, 32, 64, . . . ) the fractional loss of area diminishes.
Several methods for determining the overlap of a reader or reader relative to a control track or tracks are presented below. Each of these methods assumes that the control tracks are written simultaneously to writing the data tracks so the track to track spacing in a particular data band will be constant in spite of any wobble. In other words, the wobble in the control track(s) and the wobble in the data tracks will be identical. In any of the following embodiments, one or more control tracks are written simultaneously with the data tracks.
In a single control track embodiment, as shown in
It should be kept in mind that noise will always be present. And because this method uses the amplitude of the readback signal to determine the relative overlap of the reader, changes in the amplitude may or may not indicate a true position. The amplitude can be affected by a variety of things, not just head position. For instance, even if the reader is on the track exactly, the amplitude will still vary from such things as head-tape spacing, grain magnetization, variation in magnetic grain density, tape defects, randomness of particles in the erase band, etc.
To provide even more reliability, one embodiment uses multiple heads reading the same control track. For example, two readers can be used to read a single control track that is surrounded by two erase tracks, as in
In another embodiment, shown in
In a further embodiment, two (or more) pseudo-random bit sequences (PRBS) can be chosen instead of two sinusoids to be written on dual control tracks (adjacent or separated, one or more readers per control track). The PRBS sequences should be orthogonal to each other and rotated versions of each other. The reader signal on these two tracks may be output to two matched filters, matched to the two PRBS sequences. The ratio of the outputs of the two matched filters can be used for positioning information. This is different than the previous method, because now simple sinusoids are not merely written, but rather repeating, random-appearing patterns are present in the control tracks.
It is important that the two PRBS sequences are unique enough to be identified. For instance, they may be orthogonal. This means that the dot product of the two sequences is zero or very small. If one of the sequences is rotated by an arbitrary number of bits, meaning that a number of bits from the end of the sequence are appended to the beginning, the dot product of the rotated sequences should be very small as well.
Consider the following example. Control track 1 has a pseudo-random bit sequence and control track 2 has a different pseudo-random bit sequence. The bit sequences can cover the entire frequency domain, and preferably are optimized for a high signal to noise ratio. So from the frequency domain, the signals may be nearly indistinguishable. Accordingly, the serving is performed in the time domain. The same sequences repeat over and over in each control track. A matched filter recognizes a match in the pseudo-random bit sequence, its output goes high thereby indicating a match. A second matched filter similarly analyzes the second control track.
In another embodiment, two control tracks are written on either side of the adjacent track bundle having an erase track on both sides, as in
In yet another embodiment, control tracks also provide timing and phase information for clock recovery for the data read channels. As is well known in the art, the readback system of a storage system deciphers or decodes an incoming readback signal into 1s ands 0s, thereby translating the signal into bits that were written to the tape (read channel). The clock recovery subsystem synchronizes the clock of the readback system to the clock of the write system so the drive knows when 1s and 0s are coming in on the readback signal. Clock recovery is one of the most difficult processes in tape drive readback systems. Particularly, any loss of signal or dropout can cause the drive to lose the clock. Accordingly, any way to improve clock recovery is desirable.
Error correction is typically built in. However, all of this is downstream from the initial data read. So if the timing is lost, the effects are not known immediately and large errors are typical. So it would be desirable to ensure that the clock is properly aligned to the timing signal, as well as detect timing errors more quickly.
To assist in timing verification, the control tracks can provide frequency and phase information for the clock recovery circuits in the data read channels. A sinusoid signal of known period is preferred as it is periodic in nature, and so the velocity of the tape and thus the timing can be easily calculated.
In a further embodiment, positioning information is obtained from the phase difference of two readers RE(1), RE(3) on two control tracks WE(1), WE(2) with angled magnetic transitions. Note that additional readers RE(2), RE(4) may be provided to add robustness. The angle between the transitions tilt so that the two tracks provides phase-based position information as shown in
Each reader is preferably smaller than the width of the associated control track. As the medium passes by the readers, a pulse is generated at each written transition on the medium. When the two heads are in the middle, of the respective control tracks, the pulses arrive together. If the readers move laterally, one reader's pulse arrives sooner and the other reader's pulse arrives later. By measuring the spacing of the pulses, the positions of the readers relative to the control tracks can be determined.
To write the angled transitions, the writers are set at an angle.
The matrices described for the read signals (M) and the de-convolution of the read signals (MI) assumes that all of the readers have the same response. If the response of each reader is different, then a more complicated deconvolution algorithm may be implemented. Potential non-linearity or nonuniformity differences between readers include: (a) magnitude (amplitude); (b) asymmetry between positive and negative response; (c) differences in frequency response.
Regarding magnitude (amplitude) differences, magnitude or amplitude corrections are relatively easy to perform as long as they do not vary with time. Amplitude variations between the different readers due to their inherent differences in response are given by the matrix MA:
MA(i,i)=A(i) Equation 7
MA(i,j)=0 Equation 8
The inversion matrix (IMA) is:
IMA(i,i)=1/A(i) Equation 9
IMA(i,j)=0 Equation 10
Regarding asymmetry differences, matrices similar to MA and IMA describe asymmetric reader responses, but two values of A(i) are used:
A(i)=Ap(i) Equation 11
if R(i)>0, and
A(i)=An(i) Equation 12
In order to utilize the more complicated correction algorithms, the response of the readers must be determined. One method of determining the response of the readers is to calibrate them in a designated section of the tape. In the calibration section of the tape, all writers write the same pattern. The amplitude and asymmetry response of each reader can then be determined. Even the frequency response can be determined. It is best if the frequency response of the readers is compensated for in the hardware such as the methods employed in existing tape drives using equalizer filters to boost high frequency signals, etc. To be sure that the edge readers are calibrated properly, more writers than those writing data should be employed.
Conversion of the read signals into data bits can be accomplished using standard algorithms such as peak detect or partial response, maximum likelihood (PRML), but would be applied to the corrected read signals (CR) rather than the directly read signals (R).
A method of resolving a non-homogeneous reader profile along the reader track width is to perform an in-drive microtrack profile calibration.
where I is the identity matrix or the diagonal matrix.
The value of fn (=f(n,x)) and dn (=d(n,x)) are a function of x. The function of x has been left out of the matrix for simplicity. The value of x is determined by the “extra” (control) written track(s) by the continuously measured vales of fE (=fE(x)) and dE (=dE(x)), where the values of ffi and dn are predetermined. For actual use, the track can be divided into segments and the appropriate inversion matrices can be stored in a look-up table.
The value of f(j) and d(j) used in IM to deconvolute the signal is made by the measurement of fE and dE from the “extra” (control) track(s). For example, if all tracks have uniform response across their width, then fE=(x/WR), and dE=1−fE so all f(jr)=fE and all d(jr)=dE.
For a better determination of fE(x) and dE(x), rather than just relying on instantaneous values, an appropriate integral using an appropriate time can be used. The general form of the matrix is:
for jr=1:ndim, for jc=1:ndim, M(jr,jc)=0; end, end
M(1,1)=f(1); for jr=2:ndim, M(jr,jr)=f(jr); M(jr,jr−1)=d(jr); end
The general form of the inversion matrix is:
for jr=1:ndim, for jc=1:ndim, M(jr,jc)=0; end, end
for nc=2:jr, jc=jc−1; IM(jr,jc)=−1*(d(jc+1)/f(jc))*MI(jr,jc+1); end
The read signal given the written signal will be:
R=M*W Equation 13
The deconvoluted written (DW) tracks is given by:
DW=IM*R=IM*(M*W)=(IM*M)*W=I*W=W Equation 14
Test of Superposition Using a Real Head
As a check that the response of the readers is linear so that the signals from readers which overlap two written tracks can be deconvolved using superposition, an experiment was tried with an LTO Generation 1 (Gen 1) head. This experiment provides a first indication of the feasibility of the deconvolution techniques described herein. The LTO Gen 1 heads have SR=SW=333 μm, WR=12.6 μm and WW=26.5 μm. The LTO Gen 1 heads have 8 readers and 8 writers on two modules. The readers of the “reading” module are aligned with the writers of the “writing” module. In the experiment, the tape was first AC erased. An 8T pattern was then written for a section of tape tracks W1(8T), W2(8T), . . . W8(8T) are all separated transversely by 333 μm.
The top curve of
The best attempt to deconvolute the mixed signal is shown in
As shown, a tape supply cartridge 2220 and a take-up reel 2221 are provided to support a tape 2222. These may form part of a removable cassette and are not necessarily part of the system. Guides 2225 guide the tape 2222 across a bidirectional tape head 2226. Such bidirectional tape head 2226 is in turn coupled to a controller assembly 2228 via a compression-type MR connector cable 2230. The actuator 2232 controls position of the head 2226 relative to the tape 2222.
A tape drive, such as that illustrated in
The controller 2228 may perform any of the functionality described above including calculation of overlap, inverse matrix calculation, data recovery, etc. Alternatively, a host system may receive signals from the head (via any path) and perform the deconvolution. In a further alternative, the controller and host share duties. The controller and/or host may each contain mechanisms (e.g., logic, software modules, processors, etc.) to perform any function described herein.
TW=aRW Equation 15
The head boundaries are indicated on
hb i =i−1−e/RW, for i=0 to 7 Equation 16
Head equations assume linear inter-track interference. Each head signal is a linear combination of track signals:
h1=t1 Equation 17
h2=abs(hb1)*t1+hb2*t2 Equation 18
h3=(a−hb2)*t2+(hb3−a)*t3 Equation 19
h4=t3 Equation 20
h5=(2a−hb4)*t3+(hb5−2a)*t4 Equation 21
h6=(3a−hb5)*t4+(hb6−3a)*t5 Equation 22
h7=t5 Equation 23
Here, assume a=1.3 and e/RW=0.5. The resulting head boundary positions are calculated as:
hb1=−0.5, hb2=0.5, hb3=1.5, hb4=2.5, hb5=3.5, hb6=4.5
Plugging this information into Equations 17-23:
[add head equations with same track]
h1 = t1
h2 = 0.5*t1 + 0.5*t2
1.5*t1 + 0.5*t2 = h1 + h2
h3 = 0.8*t2 + 0.2*t3
0.5*t1 + 1.3*t2 + 0.2*t3 = h2 + h3
h4 = t3
0.8*t2 + 1.3*t3 + 0.9*t4 = h3 + h4 + h5
h5 = 0.1*t3 + 0.9*t4
0.1*t3 + 1.3*t4 + 0.6*t5 = h5 + h6
h6 = 0.4*t4 + 0.6*t5
0.4*t4 + 1.6*t5 = h6 + h7
h7 = t5
The matrix form of the equations is:
The matrix equation can be used to cancel intertrack interference. The output of the matrix equation can be input to standard 1D channels. Also note that the offtrack signal can also be derived from a servo system, and can also be updated in a detector.
The invention can take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment containing both hardware and software elements. In a preferred embodiment, the invention is implemented in software, which includes but is not limited to firmware, resident software, microcode, etc.
Furthermore, the invention can take the form of a computer program product accessible from a computer-usable or computer-readable medium providing program code for use by or in connection with a computer or any instruction execution system. For the purposes of this description, a computer-usable or computer readable medium can be any apparatus that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device.
The medium can be an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system (or apparatus or device) or a propagation medium. Examples of a computer-readable medium include a semiconductor or solid state memory, magnetic tape, a removable computer diskette, a random access memory (RAM), a read-only memory (ROM), a rigid magnetic disk and an optical disk. Current examples of optical disks include compact disk—read only memory (CD-ROM), compact disk—read/write (CD-R/W) and DVD.
A data processing system suitable for storing and/or executing program code will include at least one processor coupled directly or indirectly to memory elements through a system bus. The memory elements can include local memory employed during actual execution of the program code, bulk storage, and cache memories which provide temporary storage of at least some program code in order to reduce the number of times code must be retrieved from bulk storage during execution.
Input/output or I/O devices (including but not limited to keyboards, displays, pointing devices, etc.) can be coupled to the system either directly or through intervening I/O controllers.
Network adapters may also be coupled to the system to enable the data processing system to become coupled to other data processing systems or remote printers or storage devices through intervening private or public networks. Modems, cable modem and Ethernet cards are just a few of the currently available types of network adapters.
While various embodiments have been described above, it should be understood that they have been presented by way of example only, and not limitation. Thus, the breadth and scope of a preferred embodiment should not be limited by any of the above-described exemplary embodiments, but should be defined only in accordance with the following claims and their equivalents.
|Cited Patent||Filing date||Publication date||Applicant||Title|
|US4903150||Dec 21, 1987||Feb 20, 1990||Sharp Kabushiki Kaisha||Multitrack magnetic recording and reproducing apparatus and method of controlling tracking of multitrack magnetic heads|
|US5289328||May 26, 1992||Feb 22, 1994||George Saliba||Method and apparatus for variable density read-after-writing on magnetic tape|
|US5432652||Jun 14, 1993||Jul 11, 1995||International Business Machines Corporation||Servo and data format for magnetic tape having plural spaced-apart servo areas interleaved with data track areas having serpentine track scanning using any one of a plurality of number of concurrently accessed tracks|
|US5448430||Aug 5, 1993||Sep 5, 1995||International Business Machines Corporation||Track following servo demodulation|
|US5452152 *||Jun 6, 1991||Sep 19, 1995||Tandberg Data A/S||Tracking control on longitudinal azimuth tracks using an auxiliary read head|
|US5574602 *||Jun 14, 1993||Nov 12, 1996||International Business Machines Corporation||Processing servo signals concurrently read from plural spaced-apart servo areas for a magnetic tape having serpentine data track scanning|
|US5629813||Mar 16, 1995||May 13, 1997||International Business Machines Corporation||Initialization and calibration of magnetic tape having multiple servo areas|
|US5898534 *||May 1, 1996||Apr 27, 1999||Overland Data, Inc.||Method and system for controlling the position of a magnetic read head|
|US6104556||Nov 18, 1997||Aug 15, 2000||Seagate Technology, Inc.||Wide write disc certification process, apparatus and article of manufacture|
|US6141174 *||Aug 14, 1998||Oct 31, 2000||Quantum Corporation||Method of reading recorded information from a magnetic tape that compensates for track pitch changes|
|US6236525 *||Aug 14, 1998||May 22, 2001||Storage Technology Corporation||Tape head with pattern timing for servo writing application|
|US6330123||May 19, 2000||Dec 11, 2001||Imation Corp.||Head assembly having a single pass servo writer|
|US6496328 *||Dec 30, 1999||Dec 17, 2002||Advanced Research Corporation||Low inductance, ferrite sub-gap substrate structure for surface film magnetic recording heads|
|US6650496||May 15, 2001||Nov 18, 2003||Phs Mems||Fully integrated matrix magnetic recording head with independent control|
|1||Abbott et al., "Performance of Digital Magnetic Recording with Equalization and Offtrack Interference" IEEE Transactions on Magentics, vol. 27, No. 1, Jan. 1991.|
|2||Ahmed et al., "Track Squeeze Using Adaptive Intertrack Interference Equalization" IEEE Transactions on Magnetics, vol. 38, No. 5, Sep. 2002.|
|3||Bain et al., "Limitations to Track Following Iposed by Position Error Signal SNR Using a Multi-Tapped Magnetoresistive Servo Head" IEEE Transactions on Magnetics, vol. 35, No. 2, Mar. 1999.|
|4||Barbosa et al., "Array Heads for Estimating track Misregistration" IEEE Transactions on Magnetics, vol. 29, No. 6, Nov. 1993.|
|5||Barbosa, Lineu C., "Simultaneous Detection of Readback Signals form Interfering Magnetic Recording Tracks Using Array Heads" IEEE Transactions on Magnetics, vol. 26, No. 5, Sep. 1990.|
|6||Bellini et al., "Nonlinear Cross Talk Cancellation for High Density Optical Recording" Global Telecommunications Conference- Globecom, IEEE 1999.|
|7||Davey et al., "Two-Dimensional Coding for a Multi-Track Recording System to Comabt Inter-Track Interference" IEEE Transactions on Magentics, vol. 34, Np. 4, Jul. 1998.|
|8||Gyun Roh et al. "Ringle-Head/Single-Track Detection in Interfering Tracks" IEEE Transactions on Magnetics, vol. 38, No. 4, Jul. 2002.|
|9||Immink et al., "Signal Processing and Coding for Two-Dimensional Optical Storage" 2003.|
|10||Kato et al., "Two-Dimensional Run-Length-Limited Code and Partial Response Maximum Likelihood System with Mulut-Track Recording" WP.23, IEEE 2002.|
|11||Kumar et al., "Two-Dimensional Equalization: Theory and Applications to High Density Magnetic Recording" IEEE Transactions on Communications, vol. 42, No. 2/3/4, Feb./Mar./Apr. 1994.|
|12||Kurtas et al., Reduced Complexity Maximum Likelihood Sequence Estimation for Multitrack High-Density magnetic Recording Channels IEEE Transactions on Magnetics, vol. 35, No. 4, Jul. 1999.|
|13||Office Action Summary from Appl. No. 11/216,615 which was mailed on Aug. 7, 2006.|
|14||Singla et al, "Iteratice Decoding and Equalization for 2-d Recording Channels" IEEE Transactions on Magnetics, vol. 38, No. 5, Sep. 2002.|
|15||Soljanin et al., "Multihead Detection for Multitrack Recording Channels" IEEE Transactions on Information Theory, vol. 44, No. 7, Nov. 1998.|
|16||Steele II et al., "Multi-Tapped Magentoresistive Heads for Magnetic Tape tracking Servo" IEEE Transactions on Magnetics, vol. 34, No. 4, Jul. 1998.|
|17||U.S. Appl. No. 11/216,615, filed Aug. 30, 2005.|
|18||Vea et al, "Multichannel Equalization For High Track Densisty Magnetic Recording" IEEE, 1994.|
|19||Vea et al., "Magnetic Recording Channel Model with Intertrack Interference" DSSC, Depart. Elect. And Comp. Eng., Carnegie Mellon University, Pittsburgh, PA, 15213, IEEE 1994.|
|20||Voois et al, "A Decision Feedback Equalizer for Multiple-Head Digital Magnetic Recording" Information Systems Laboratory, Stanford University, Stanford, CA. 94305, IEEE 1991.|
|21||Voois et al, "Multichannel Digital Magnetic Recording" Information Systems Laboratory, Stanford University, Stanford, CA. 94305, IEEE 1992.|
|22||Voois et al., "Multichannel Signal Processing for Multiple-Head Digital Magnetic Recording" IEEE Transactions on Magnetics, vol. 30, No. 6, Nov. 1994.|
|23||Voois et al., "Upper Bound on Achievable Storage Density: A Two-Dimentional Approach" IEEE Transactions on Magnetics, vol. 33, No. 1, Jan 1997.|
|Citing Patent||Filing date||Publication date||Applicant||Title|
|US7428120 *||Mar 30, 2007||Sep 23, 2008||International Business Machines Corporation||System and method for deconvolution of multiple data tracks|
|US7724459||Sep 5, 2008||May 25, 2010||International Business Machines Corporation||Method for reading data from a magnetic recordign tape|
|US8379498||Jul 19, 2011||Feb 19, 2013||Lsi Corporation||Systems and methods for track to track phase alignment|
|US8681439||Jul 19, 2011||Mar 25, 2014||Lsi Corporation||Systems and methods for handling sector gaps in inter-track interference compensation|
|US8749912||Jul 20, 2010||Jun 10, 2014||International Business Machines Corporation||Method and apparatus for bipolar servo marks writing with self DC-erase|
|US8773794||Jul 19, 2011||Jul 8, 2014||Lsi Corporation||Systems and methods for block-wise inter-track interference compensation|
|US8804260||Jul 19, 2011||Aug 12, 2014||Lsi Corporation||Systems and methods for inter-track interference compensation|
|US8810941||Jul 10, 2012||Aug 19, 2014||HGST Netherlands B.V.||System, method and apparatus for selection of reserved area data tracks|
|US8830613||Aug 22, 2012||Sep 9, 2014||Lsi Corporation||Storage media inter-track interference cancellation|
|US8879180||Dec 12, 2012||Nov 4, 2014||HGST Netherlands B.V.||System, method and apparatus for data track usage sequence to reduce adjacent track interference effect|
|US8879182||Aug 22, 2012||Nov 4, 2014||Lsi Corporation||Storage media inter-track interference cancellation|
|US9007712 *||Dec 16, 2013||Apr 14, 2015||International Business Machines Corporation||Backward compatible head for quasi-static tilted reading and/or recording|
|US9055711||Jul 16, 2012||Jun 9, 2015||HGST Netherlands B.V.||System and method for maintaining data integrity on a storage medium|
|US9117470||Jul 17, 2014||Aug 25, 2015||International Business Machines Corporation||Write delay to de-skew data in read while write function for tape storage devices|
|US9129614||May 1, 2013||Sep 8, 2015||International Business Machines Corporation||Magnetic head having canted arrays|
|US9208809||May 1, 2013||Dec 8, 2015||International Business Machines Corporation||Magnetic head and system having offset arrays|
|US9214164||Sep 16, 2013||Dec 15, 2015||International Business Machines Corporation||Miniskirt tape head having quasi-statically tilted transducer arrays|
|US9218838||Dec 12, 2013||Dec 22, 2015||International Business Machines Corporation||Quasi-statically tilted head having offset reader/writer transducer pairs|
|US9251825||Mar 11, 2015||Feb 2, 2016||Globalfoundries Inc.||Backward compatible head for quasi-static tilted reading and/or recording|
|US9263087||Oct 29, 2014||Feb 16, 2016||International Business Machines Corporation||Quasi-statically tilted head having dilated transducer pitch|
|US9299365||Sep 21, 2015||Mar 29, 2016||International Business Machines Corporation||Miniskirt tape head having quasi-statically tilted transducer arrays|
|US9311937 *||Jul 15, 2014||Apr 12, 2016||Marvell World Trade Ltd.||Systems and methods for calibrating read and write operations in two dimensional magnetic recording|
|US9318138||Dec 17, 2014||Apr 19, 2016||International Business Machines Corporation||Write delay to de-skew data in read while write function for tape storage devices|
|US9330686||Sep 21, 2015||May 3, 2016||International Business Machines Corporation||Miniskirt tape head having quasi-statically tilted transducer arrays|
|US9355665||Nov 5, 2015||May 31, 2016||International Business Machines Corporation||Quasi-statically tilted head having offset reader/writer transducer pairs|
|US20070165326 *||Mar 30, 2007||Jul 19, 2007||International Business Machines Corporation||System and method for deconvolution of multiple data tracks|
|US20080273258 *||Jul 24, 2008||Nov 6, 2008||David Berman||System and method for deconvolution of multiple data tracks|
|US20080316632 *||Sep 5, 2008||Dec 25, 2008||Czarnecki W Stanley||Method for reading data from a magnetic recording tape|
|US20150022916 *||Jul 15, 2014||Jan 22, 2015||Marvell World Trade Ltd.||Systems and methods for calibrating read and write operations in two dimensional magnetic recording|
|U.S. Classification||360/77.01, G9B/5.203, G9B/5.008|
|Cooperative Classification||G11B5/00826, G11B5/584|
|European Classification||G11B5/008T2F2, G11B5/584|
|Nov 18, 2005||AS||Assignment|
Owner name: INTERNATIONAL BUSINESS MACHINES CORPORATION, NEW Y
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:BERMAN, DAVID;CZARNECKI, W. STANLEY;IBEN, ICKO E.T.;AND OTHERS;REEL/FRAME:017044/0078;SIGNING DATES FROM 20050824 TO 20050829
|Mar 14, 2011||REMI||Maintenance fee reminder mailed|
|Aug 7, 2011||LAPS||Lapse for failure to pay maintenance fees|
|Sep 27, 2011||FP||Expired due to failure to pay maintenance fee|
Effective date: 20110807